This study proposes to compare the efficacy of Problem Solving Therapy (PST) to that of Supportive Therapy (ST) in non-demented elderly patients with major depression and cognitive impairment. Among them, we will focus on patients with major depression and executive dysfunction because this combination of symptoms (as defined in our preliminary studies) is prevalent, debilitating, and responds poorly to treatment with antidepressant agents. Therefore, we find it compelling to identify an effective treatment for these patients, who would otherwise remain depressed, debilitated, and demoralized during the last years of their lives. We selected PST because it can address depression as well as deficits in problem solving skills that impact on the patients' ability to negotiate their environment and contribute to their depression and disability. Moreover we now have empirical evidence suggesting that PST can reduce depressive symptoms and disability in cognitively unimpaired depressed elderly patients as well as elderly patients with major depression and executive dysfunction. The subjects will be 240 (120 from each Center) elderly (>64 years) patients with non-psychotic, unipolar major depression and executive dysfunction and will be randomly assigned to receive 12 sessions of PST or ST. The study is designed to test the hypotheses that the PST is more effective than ST in reducing depressive symptoms and disability. Furthermore, we hypothesize that these effects are mediated by improvement in generation of alternative solutions, decision making, and solution implementation. While we are aware of the methodological problems and confounds posed by studying a "sick and old population", we bring to this project two groups with complementary expertise in treatment studies and psychotherapy research, and experience in directing multicenter studies. Moreover, the project will be supported by the structures of the Cornell Intervention Research Center (IRC), whose principal objective is to develop treatment interventions targeting specific clinical and biological characteristics of geriatric depression. Accordingly, we are well positioned to meet the challenges inherent in this difficult but important area of research.